Cell-based therapeutics hold promise to revolutionize medicine, being applicable to numerous disease states. Due to the complex nature of cell-based medicine, its transition from proof-of-concept to clinical use has been slow. A major factor limiting the translation of cell-based therapeutics is the difficulty in tracking the fate of the cells in vivo. Positron emission tomography (PET) is an imaging technology that allows in vivo monitoring in real time and in a non-invasive manner. The signal measured in PET comes from radioactive isotopes, such a fluorine-18, that undergoes positron emission. Thus, a molecule containing such an isotope - called a PET probe - that specifically accumulates inside the therapeutic cells would reveal their location, number, half-life, etc. An additional factor limiting the advancement of therapeutic cells to the clinic has been the concern of unforeseen side effects from this novel approach. A safety mechanism that would allow for the elimination of the cells would greatly diminish this concern. The long-term goal of this proposal is to develop a dual-purpose system that would allow, one, for the real-time tracking of therapeutic cells using PET, and two, for the elimination of the cells if needed. Specific accumulation of the PET probe inside the therapeutic cells is achieved when the probe is phosphorylated in such cells. Phosphorylation traps the molecule within the cell due since charged molecules cannot traverse cell membranes. We propose to insert the gene of a human nucleoside kinase to the therapeutic cells, with the requirement that the kinase would impart unique activity to cells that express it. Such a unique enzymatic activity of a PET reporter enzyme would allow to preferentially phosphorylate novel PET probes in the therapeutic cells. The field currently relies on non-human enzymes, such as HSV1-TK, for that unique enzymatic activity. However, as a PET reporter enzyme HSV1-TK has several drawbacks, foremost being immunogenic. To circumvent this, we will employ a modified version of human thymidine kinase 2 (TK2) as the source for the unique enzymatic activity. To differentiate this enzyme from endogenous TK2, we will perform enzyme engineering to supply us with TK2 variants with an activity profile different to that of wild type TK2, but that still do not elicit an immune response. The immediate goals of this proposal are to study the determinants of substrate specificity of TK2 by solving crystal structures of the kinase in complex with substrates (Aim 1), and to exploit this understanding for the design of TK2 variants with unique activity towards novel PET probes (Aim 2). We will test the ability to track cells expressing our TK2 variants in mouse models, and compare the TK2-system to the standard in the field using HSV1-TK. The results of this research, a non-immunogenic PET reporter enzyme with optimized activity with PET probes, will allow the in vivo real time tracking of therapeutic cells. This would have a dramatic impact on the transition of therapeutic cell approaches to the clinic.